- Email: [email protected]
Value of host range, morphological, and genetic characteristics within the Entomophthora muscae species complex
mycological research 110 (2006) 941–950
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/mycres
Value of host range, morphological, and genetic characteristics within the Entomophthora muscae species complex Annette Bruun JENSEN*, Lene THOMSEN, Jørgen EILENBERG Department of Ecology, The Royal Veterinary and Agricultural University (KVL), Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark
article info
abstract
Article history:
Entomopthora muscae sensu lato is a complex of morphologically similar fungal species path-
Received 25 November 2005
ogenic to evolutionarily advanced flies (Cyclorrhapha). To reach an operational species
Received in revised form
definition and recognition of species within this complex, the values of host range,
25 April 2006
morphological and genetic characteristics are reconsidered. Within the E. muscae species
Accepted 1 June 2006
complex morphological and nuclear characteristics of the primary conidia are taxonomi-
Published online 14 August 2006
cally important. In this study we compared the dimensions and nuclear numbers of the
Corresponding Editor:
primary conidia of isolates from their original (natural) hosts and after being transferred
Richard A. Humber
to alternative hosts (cross-transmission) in order to check the stability of these characteristics. The conidial characteristics change substantially when produced in alternative host
Keywords:
species, but their overall range in variability still fit within the traditional morphological
Anthomyiidae
species circumscriptions. The phylogenetic analyses of the ITS II and LSU rRNA gene
Diptera
sequences, revealed three distinct lineages within the complex: E. schizophorae, E. muscae
Ecology
and E. syrphi. Within each of these lineages sequence divergence was seen between isolates
Entomophthorales
originating from different host species. Our studies on the physiological host range showed
Host range
that several isolates were able to infect alternative dipteran species. Musca domestica was
Insect pathogenic fungi
a particularly good receptor. The ecological host range of any individual isolate seems,
Morphology
however, to be limited to one host species evidenced by the occurrence of distinct geno-
Muscidae
types within each natural infected host species shown by RAPD. The high host specificity
Phylogeny
of these fungi emphasizes the importance of identifying the host taxon at species level in
Taxonomy
the recognition of Entomophthora species. We recommend that morphological characteris-
Zygomycota
tics of fungal structures and host taxon, together with molecular data, serve as criteria for species determination in future studies on members of the E. muscae complex. ª 2006 The British Mycological Society. Published by Elsevier Ltd. All rights reserved.
Introduction Fungi from the Entomophthora muscae species complex are important for the natural population regulation of many dipteran species. Epizootics have often been reported also among pest flies, and the fungi possess a potential to be used in biocontrol
(Klingen et al. 2000; Mullens et al. 1987; Steinkraus et al. 1993; Watson & Petersen 1993). However, the evaluation of this potential requires a more complete understanding of the ecological interaction between host species and their pathogens. In such ecological studies the circumscriptions of individual fungal taxa become a central issue. Traditionally, species were
* Corresponding author. Tel.: þ45 35 28 26 82; fax: þ45 35 28 26 70 E-mail address: [email protected]. 0953-7562/$ – see front matter ª 2006 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.mycres.2006.06.003
942
described based on recognizable morphological features and as such E. muscae from Musca domestica was described by Cohn in 1855. Subsequently, fungal specimens from flies of derived clades (Cyclorrhapha) with E. muscae-like conidia were designated as E. muscae (MacLeod et al. 1976). In the 1980s cytological features were considered important for the taxonomy of the order Entomophthorales (Ben-Ze’ev & Kenneth 1982; Humber 1981, 1989; Remaudie`re & Keller 1980), in particular the contents of heterochromatin, number and size of the nuclei. Inspired by these new criteria, Keller (1984) suggested that E. muscae s.l. is a complex of species separated by conidial size, number and size of the nuclei. At present six species are recognized and described morphologically: E. ferdinandi, E. grandis, E. muscae s.str., E. scatophagae, E. schizophorae, and E. syrphi (Keller 2002). Some of the species, e.g., E. schizophorae, E. ferdinandi, and E. grandis, have distinct conidial morphology whereas others such as E. muscae s.str. and E. scatophagae are virtually indistinguishable using conidial characteristicistics. The two latter species can only be differentiated morphologically by the shape of their hyphal bodies, which are reported to be consistently spherical in E. muscae s.str. (Keller et al.1999). Whether this morphological variation is stable or due to phenotypic modifications associated with the environmental condition (the host species) remains unknown. Host range has been used to elucidate entomophthoralean species complexes. Based on transmission experiments, RFLP, and hybridization of a DNA probe, Entomophaga grylli has been divided into pathotypes according to host subfamily (Ramoska et al. 1988). Several laboratory transmission experiments of E. muscae s.l. have been performed within a range of Diptera (Baird 1957; Eilenberg et al. 1987; Kramer & Steinkraus 1981; Mullens 1989; Steenberg et al. 2001; Steinkraus & Kramer 1987), but no attempt to group E. muscae into pathotypes or groups based on their host range has been made. However, E. scatophagae is believed to have a restricted host range and is only described from a single host species; E. grandis and E. syrphi have only been reported from larger and smaller syrphid flies, respectively (Keller 2002). A recent study of the intraspecific variation within E. muscae s.str. showed that several genotypes occur and that each type was restricted to a single host species (Jensen et al. 2001). This indicates a high degree of host specificity in the E. muscae species complex similar to that observed in the E. grylli complex. Phylogenetic relationships of the order Entomophthorales have been inferred from the SSU rRNA gene sequences (Jensen et al. 1998; Nagahama et al. 1995); but until now no sequence analyses have been accomplished to resolve species complexes within Entomophthorales. Nevertheless, PCR-RFLP of the ITS II and the 50 end of the LSU rRNA gene support the separate origin of E. muscae, E. schizophorae, and E. syrphi (Jensen & Eilenberg 2001). The aim of this study was to evaluate the value of host range, morphological, and genetic characteristics for operational species recognition of species within the E. muscae complex. Studies were made to elucidate specific topics: (1) the plasticity of certain morphological characteristics was explored, (2) the physiological host range among certain host taxa was examined, and (3) the phylogeny was assessed by sequence analyses of the ITS II and part of the LSU rRNA gene.
A. B. Jensen et al.
Materials and methods Field samplings Approximately 8000 live flies (Table 1) were collected by sweep-net in Denmark during 1999 (June–September) and 2000 (May–October) from the following habitats: cabbage and rape fields, border vegetation, forests, stables, and houses. Live flies were transferred individually to 30 ml cups containing 5 ml 1.5 % water-agar as described by Eilenberg & Philipsen (1988). Condensation in the cup was avoided by cutting a hole in the lid and covering it with gauze. Food (a paste of 9 parts sugar, 1 part dry yeast, and little water) was applied directly onto the gauze. Predatory flies (predominantly Coenosia tigrina and Scatophaga stercoraria) were also supplied with live houseflies every third day, and hover flies (Syrphidae) were provided with flowers or pollen. The flies were incubated at room temperature and exposed to natural day length. They were examined for death and external signs of fungal infection every day in a 14 d period. Conidia were collected on glass slides in a humid chamber from all the cadavers and stored dry until morphological examination. Immediately after the conidial collection, well-sporulating cadavers were either used in the host range transmission experiments or were used to obtain in vitro isolates to ensure maximum inoculum levels.
Morphological characteristics The morphology of primary conidia discharged from the original hosts was compared with the morphology of the same in vivo-isolate after cross-transmission to an alternative host. The dimensions of 20 primary conidia per isolate were measured mounted in lactophenol-cotton blue (0.01 % cotton blue) using an Olympus Provis microscope supplemented with an Oly-Lite computer-based system for morphometrics at 400 magnification. The number of nuclei of 20 primary conidia was counted mounted in DAPI (4,6-diamidino-2-phenylindole, 5 mg ml1) using a Zeiss Axiophot epifluorescence microscope (filter 0.5; violet, excitation 395–440) at 400 magnification.
Table 1 – Number of collected flies and number of flies with external sign of Entomophthora muscae s.l. Host taxa
Anthomyiidae Anthomyiidae sp. Delia radicum Calliphoridae Pollenia angustigena/rudis Muscidae Coenosia tigrina Musca domestica Scatophagidae Scatophaga stercoraria Syrphidae Syrphidae sp.1
Approx. number of collected flies
Number of Entomophthora infected flies
350 1100
25 250
900
90
300 700
20 100
1650
90
1000
90
1 From the genera Melanostoma and Platycheirus.
Entomophthora muscae species complex
943
Standard LSD tests (PROC GLM; SAS Institute 1999) were used to test if the interaction of host and isolate had significant influence on the length, width, and number of nuclei in the primary conidia of Entomophthora muscae and E. schizophorae. Within each isolate comparison of mean values were made using the Tukey test.
this study were stored in the Fungal Insect Pathogenic Culture Collection at the Royal Veterinary and Agricultural University, Copenhagen, by cryopreservation at 80 C according to Lo´pez Lastra et al. (2001) and are deposited in ARSEF (USDAARS, Ithaca, NY). All the isolates used are listed in Tables 2 and 3.
In vitro isolation
Physiological host range
In vitro isolation of the Entomophthora species was done as in Jensen et al. (2001). Fresh, well-sporulating cadavers were placed on a small sterile inverted Petri dish lid (3.5 cm diam) and the base was put on top. The wings of the large flies, Musca domestica, Myospila meditabunda, Pollenia angustigena, P. rudis, and Scatophaga stercoraria were cut off in order to avoid contamination by wings hitting the Petri dish base. Sporulation was allowed for 30 min and the arrangement was placed in a humid chamber in order to enhance sporulation. In a sterile hood 1 ml GLEN (Beauvais & Latge´, 1988) including 5 % fetal bovine serum was poured onto the discharged conidia. A new sterile lid was used and the petri dish was sealed with Parafilm and incubated dark at 23 C. When growth was observed, the in vitro culture was transferred to 50 ml cell culture flasks with 5 ml GLEN for further growth. The isolates used in
Laboratory transmission experiments were conducted to obtain information on physiological host range. Two types of transmission experiments were performed: in small cup experiments, one to five live flies (receptors) and a wellsporulating cadaver (donor) were placed together in a 30 ml cup containing 5 ml 1.5 % water-agar. Cups with live flies and sporulating cadavers were maintained at 18 C and received a photoperiod of 16:8 (light:dark) in 24 h for fungal inoculation. Alternatively, in caged experiments, ten to 20 live flies (receptors) were placed in a cylindrical cardboard container (9 cm diam, 5 cm high). The cardboard containers had a Petri dish base (9 cm diam) as the bottom and a 2-mm mesh top as a cover. Two to five well-sporulating cadavers (donors) were placed on the mesh top and a Petri dish lid (9 cm diam) was put on top of the sporulating cadavers. The
Table 2 – Insect host and geographic origin of the in vitro isolates used for RAPD Isolatea KVL 00-56 KVL 00-57b KVL 00-58 KVL 00-59 KVL 00-60 (ARSEF 6704) KVL 00-61 (ARSEF 6705) KVL 00-62 KVL 00-63 KVL 00-64 KVL 00-65 KVL 00-66 KVL 00-67 NCRI 1-99 (ARSEF 6274) ARSEF 2668 KVL 99-102 (ARSEF 6815)b KVL 99-103 KVL 99-96 KVL 99-85(ARSEF 6811) KVL 99-13 (ARSEF 6132)bc KVL 99-23 (ARSEF 6141) KVL 99-110 (ARSEF 6813) ARSEF 2542 KVL 99-63 KVL 99-70 KVL 99-90 KVL 99-92 (ARSEF 6918) KVL 99-87 NCRI 2-99
Fungal species
Host species (host family)
Collection site
Year
Entomophthora scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae Entomophthora muscae E. muscae E. muscae E. muscae E. muscae E. muscae E. muscae s.str. E. muscae s.str. E. muscae s.str. E. muscae s.str. E. muscae s.str. E. muscae s.str. E. ferdinandi E. ferdinandi E. ferdinandi E. ferdinandi
Scatophaga stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) Delia radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) Musca domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) Coenosia tigrina (Muscidae) C. tigrina (Muscidae) Pegoplata infirma (Anthomyiidae) Botanophila fugax (Anthomyiidae)
Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Ski, Norway Svanholm, Denmark Sengeløse, Denmark Ørslev, Denmark Sengeløse, Denmark Hegnstrup, Denmark Kirke Sa˚by, Denmark Skævinge, Denmark Knardrup, Denmark Ribe, Denmark Sverkildstup, Denmark Sverkildstup, Denmark Ørslev, Denmark Ørslev, Denmark Hegnstrup, Denmark Ski, Norway
2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 1999 1986 1999 1999 1999 1999 1999 1999 1999 1986 1999 1999 1999 1999 1999 1999
a KVL, Danish abbreviation for the Royal Veterinary and Agricultural University, Copenhagen; ARSEF, ARS Collection of Entomopathogenic Fungi, Ithaca; NCRI, Norwegian Crop Research Institute. b Isolates also used for the sequence analyses. c In vitro isolate made from in vivo cultures, that was also used for the transmission experiments.
KVL, Danish abbreviation for the Royal Veterinary and Agricultural University, Copenhagen; ARSEF, ARS Collection of Entomopathogenic Fungi, Ithaca. Isolates also used for the RAPD analysis. In vitro isolate made from in vivo cultures, that was also used for the transmission experiments. Genbank accession numbers for the sequences. a b c d
1999 1999 2000 1996 2000 2000 1998 Kirke Sa˚by, Denmark Sengeløse, Denmark Tuse Næs, Denmark Roskilde, Denmark Hundested, Denmark Hundested, Denmark Gribskov Denmark Musca domestica (Muscidae) Delia radicum (Anthomyiidae) Scatophaga stercoraria (Scatophagidae) M. domestica (Muscidae) Pollenia rudis (Calliphoridae) M. meditabunda (Muscidae) Melanostoma mellinum (Syrphidae) KVL KVL KVL KVL KVL KVL KVL
99-13 (ARSEF 6132)b,c 99-102 (ARSEF 6815)b 00-57 (ARSEF 6704)b 99-18 (ARSEF 6137) 00-93 (ARSEF 6817) 00-92 (ARSEF 6701) 98-19 (ARSEF 5955)
Entomophthora muscae s.str. E. muscae E. scatophagae E. schizophorae E. schizophorae Entomophthora sp. E. syrphi s.str.
16–18 14–16 16–18 5–7 5–7 20–23 17–21
DQ481217/DQ481224 DQ481218/DQ481225 DQ481219/DQ481226 DQ481220/DQ481227 DQ481221/DQ481228 DQ481222/DQ481229 DQ481223/DQ481230
Year Host (family) Isolatea
Species
Nuclei number
Accession no.d ITSII/LSU
Collection site
A. B. Jensen et al.
Table 3 – Insect host and geographic origin of the isolates used for sequencing of the ITS II and the first part of the LSU rRNA gene
944
cardboard assemblies with live flies and sporulating cadavers were maintained 18 C and received a photoperiod of 16:8 (light:dark) in 24 h for fungal inoculation. The cage type was only performed with donors from two in vivo cultures of Entomophthora muscae s. str (DK3) (Keller et al. 1999) and E. schizophorae (DPIL 150896) (Kalsbeek et al. 2001) and with Musca domestica and Delia radicum as receptors. The two in vivo cultures were maintained by continuous passages in M. domestica as described by Kramer and Steinkraus (1981), and both had M. domestica as original host. After the inoculation, the flies were placed individually in 30 ml cups containing water agar and covered with gauze and a lid with a hole. The sugar–yeast paste described above was provided as food directly on the gauze. The flies in the cups were incubated at 18 C with a photoperiod of 16:8 (light:dark) for three weeks, and were examined daily for death and external signs of fungal infection. Conidia from the cadavers were collected and stored dry until examination. M. domestica originated from a laboratory stock and was supplied as pupae by the Danish Institute of Agricultural Sciences, Department of Integrated Pest Management, Danish Pest Infestation Laboratory Group (DPIL). The other receptor flies originated from the field but were used only after a quarantine period of at least two weeks to exclude contamination from natural infections. Thus the field collected flies were old and therefore those that died during the transmission experiments without external signs of Entomophthora ssp. infections were excluded. Logistic regression was used to analyse the effects of fungal species and the receptor fly on the percentage of infected flies (Proc GENMOD; SAS Institute 1999) with a binomial distribution and logit as link functions. Over-dispersion was taken into account (Collet 1991). To test the effect of receptor fly, all receptor flies were categorized into three groups: (1) the receptor fly was of the same species as the donor fly, (2) the receptor fly was M. domestica, and (3) the receptor fly was neither the same fly species as the donor fly or M. domistica.
Genetic characteristics DNA extraction and amplification of ITS II and the LSU rRNA gene of seven in vitro isolates (Table 2) were conducted as described by Jensen et al. (2001). For the amplification of the ITS II the primers nu-5.8 S-50 (Jensen & Eilenberg 2001) and ITS 4 (White et al. 1990) were used and for the amplification of the LSU the primers nu-LSU 0018-50 (Jensen & Eilenberg 2001) and nu-LSU 0805-30 (Kjøller & Rosendahl 2000) were used. The PCR conditions were initial denaturation for 5 min at 96 C, followed by 35 cycles with denaturation for 1 min at 96 C, annealing for 1 min at 62 C (ITS II) or 55 C (LSU), extension for 1 min at 72 C and a final extension for 10 min at 71 C. The PCR reactions were carried out in a 25 ml volume with 250 mM of each dNTP, 0.8 mM of each primer, 2.5 mM MgCl2, 1 buffer (10 mM Tris–HCl, pH 8.8 at 25 C, 50 mM KCl, 0.1 % Triton X-100), 1 unit DyNazyme II (Finnzymes, Espoo, Finland) and 1 ml extracted DNA, diluted at 1:10 or 1:100. Before sequencing, the PCR products were purified with the GFXtm PCR DNA and Gel Purification Kit (Amersham Bioscience, Uppsala, Sweden). Both strands of the ITS II and LSU were sequenced by GATC GmbH. To get the full sequence of the ITS II
Entomophthora muscae species complex
four internal primers ITS II 500-50 (5’-ACGTTGGCAACGAC WAGTAA-3’), ITS II 500-3’ (5’-TTACTWGTCGTTGCCAACGT3’), ITS II 1000-5’ (5’-TTGTTGAAWCTTTTCTGTCGC-3’) and ITS II 1000-30 (50 -CGACAGAAAAGWTTCAACAAT-30 ) were used. The chromatograms were checked using Sequencer 3.1 (Gene Codes Corporations, Ann Arbor, MI). In addition the LSU sequence of Entomophaga aulicae (Genbank accession no: U35394) was included, but the E. aulicae ITS II sequence was too divergent to be aligned with the sequences generated in this study. Following initial alignment with CLUSTAL V 1.60 (Higgins et al. 1992) the alignments were adjusted manually. MP analyses were performed with the PAUP 4.0b6 (Swofford 1998) on individual data sets (ITS II and LSU) and combined data set (ITS II þ LSU). The branch-and-bound search option was used to find exact solutions. All characteristics were equally weighted, and invariant characteristics were ignored. In the LSU analysis E. aulicae was chosen as an outgroup. Supports for internal branches were assessed by 1000 BS replications, and the partition-homogeneity test was used to evaluate the concordance between two individual data sets. NJ analyses were performed with the software Treecon for Windows (Van de Peer & De Wachter 1994) using the JukesCantor evolutionary model. Supports for internal branches were assessed by 1000 BS replications. RAPD was performed with the 10 commercial RAPD primers (OPA01, OPA02, OPA05, OPA06, OPA07, OPA09, OPA10, OPA15, OPA17,OPA20; Operon Technologies, Alameda, CA) on 11 Entomophthora scatophagae, six E. muscae, six E. muscae s.str., and four E. ferdinandi in vitro isolates originating from various host species (Table 2) as described in Jensen et al. (2001).
Results Morphological characteristics The primary conidia all were campanulate and multinucleate irrespective of host origin. The overall analysis of the primary conidia from the cross-transmission experiments showed significant effects on the interaction of host species and isolate on length (F4, 266 ¼ 75.80; P < 0.0001), diameter (F4, 266 ¼ 49.16; P < 0.0001), and number of nuclei (F4, 266 ¼ 3.83; P ¼ 0.0048) for Entomophthora muscae and on length (F1, 133 ¼ 19.19; P < 0.0001) and diameter (F1, 133 ¼ 89.78; P < 0.0001) for E. schizophorae. The number of nuclei in E. schizophorae varied a little, but was not significantly different irrespective of host species (F1, 133 ¼ 1.29; P ¼ 0.2587). No clear pattern in size or number of nuclei after transmission to alternative hosts was detected, except for Pollenia angustigena/P. rudis where conidia from the original host were slightly larger and contained more nuclei than conidia of the same isolates produced from other dipteran hosts (Table 4). However, the changes in morphology after transmission to alternative hosts were never so dramatic that they influenced the classical morphological species identification.
Physiological host range Laboratory transmission of Entomophthora muscae s.l. between different host species was possible, but with variable success
945
(Table 5). The mortality attributable to the entomophthoralean fungi was significantly affected by the receptor fly (c2 ¼ 15.89; df ¼ 2; P ¼ 0.0004). Having Musca domestica as receptor fly compared with having the same receptor as donor fly species did not show significantly increased infectivity, as tested using Pair-wise comparisons using least squares means to test the effects of the species of receptor fly [Pr > |t| for H0: LSMean (category i) ¼ LSMean (category j)] (c2 ¼ 0.55; df ¼ 1; P ¼ 0.4593), whereas all other comparisons were significantly different (P 0.0089). Whereas M. domestica was an excellent receptor for all isolates tested (except for those from E. syrphi), Coenosia tigrina proved difficult to infect in our study conditions, even with E. muscae from naturally infected C. tigrina. Interestingly we observed a decrease in the number of conidiophores, as only one or two narrow abdominal bands of conidiophores appeared in several of the alternative hosts, thus resulting in moderate to restricted sporulation.
Genetic characteristics Approximately 1440 bp of the ITS II and 900 bp of the 50 end of the LSU rRNA gene were sequenced from the seven isolates listed in Table 3. The ITS alignment had 306 informative sites, whereas the LSU alignment, including Entomophthora aulicae, had 56 informative sites. Both regions produced trees with similar topology using maximum parsimony (Fig 1A–B). The parsimony-based partition homogeneity test found no support for incongruence between these two data sets (P ¼ 1.000), and thus the two data sets were combined. The tree topology of the combined data sets was similar to the tree topology of the individual data sets (Fig 1C), as were also the topology of the NJ trees (data not shown). Three major lineages of the E. muscae species complex were resolved, each of which centres on a single species; E. muscae, E. schizophorae, and E. syrphi. Each of the lineages was supported by BS values above 92 %. The E. syrphi and E. muscae lineages clustered together on a branch supported by a BS value of 74 % in the LSU analysis. In this study E. scatophagae was isolated in vitro for the first time. All 11 E. scatophagae isolates obtained had similar RAPD profiles with the ten RAPD primers used, but were different from E. muscae isolates originating from other host species (Fig 2).
Discussion Host range has long been regarded to be an important characteristic in the circumscription of Entomophthora species. Physiological host range can be reflected in laboratory transmission experiments, and our study showed that receptor flies from other (unnatural) host taxa generally were less susceptible than the original (natural) host flies, which is concordant to other transmission experiments performed (Baird 1957; Eilenberg et al. 1987; Kramer & Steinkraus 1981; Mullens 1989; Steenberg et al. 2001; Steinkraus & Kramer 1987). Those specimens that got infected often showed a decreased sporulation, which would consequently decrease further disease transmission. However, Musca domestica was equally susceptible to E. muscae s.l., compared with the exposed original host species.
946
A. B. Jensen et al.
Table 4 – Dimensions and numbers of nuclei in primary conidia of Entomophthora muscae and E. schizophorae before and after transmission to alternative host species Transmission experiments
Fungal species
Host species
Mean no. of nuclei
Mean length (mm)
Mean width (mm)
1 1
Entomophthora muscae E. muscae
Scatophaga stercorariaa Musca domestica
17.4 a 15.6 b
27.2 a 28.2 b
21.8 a 22.4 a
2 2
E. muscae E. muscae
Delia radicuma M. domestica
17.1 a 16.1 b
29.2 b 25.8 a
22.6 b 20.1 a
3 3 3
E. muscae E. muscae E. muscae
D. radicuma M. domestica M. domestica / D. radicumb
16.6 a 15.6 b 15.6 b
24.6 a 27.1 c 25.6 b
18.5 a 22.3 c 20.5 b
4 4
E. muscae E. muscae
D. radicuma M. domestica
14.8 a 15.5 a
23.7 a 25.7 b
18.5 a 19.4 b
5 5
E. muscae E. muscae
D. radicuma M. domestica
15.3 a 16.0 a
24.3 a 28.2 b
18.5 a 23.2 b
6 6 6
E. muscae s.str.d E. muscae s.str.d E. muscae s.str.d
M. domesticaa D. radicum Pollenia angustigena/rudisc
16.7 b 17.0 b 19.4 a
29.7 b 25.4 a 31.4 c
22.8 b 20.7 a 25.1 c
7 7 7
E. schizophoraee E. schizophoraee E. schizophoraee
M. domesticaa D. radicum P. angustigena/rudisc
5.5 a 5.5 a 5.6 a
22.6 a 23.2 a 23.9 b
17.9 a 18.0 ab 18.5 b
8 8
E. schizophorae E. schizophorae
Coenosia tigrinaa M. domestica
5.6 a 5.5 a
20.4 a 20.5 a
14.5 a 15.1 b
9 9
E. schizophorae E. schizophorae
P. angustigenaa M. domestica
6.1 a 5.5 a
23.8 a 20.7 b
19.3 b 15.2 a
All the cadavers used in the transmission experiments came from the field sampling, except the for the Entomophthora muscae s.str. and E. schizophorae with M. domestica as original host. Within each transmission experiment comparison of the mean values were made using a Tukey test and mean values followed by different letters are significantly different (P < 0.05). a The original host species. b This isolate was transmitted from Delia radicum to Musca domestica and then back to D. radicum again. c A mixed population of Pollenia angustigena and P. rudis. d The cadavers came from the in vivo culture DK3. e The cadavers came from the in vivo culture DPIL 150896.
Nevertheless, only a single genotype of E. muscae s.str. was previously documented from Danish collections of M. domestica (Jensen et al. 2001) sampled at different locations and years, so the high susceptibility of M. domestica to several E. muscae genotypes seems not to be reflected in natural epizootics. Within the E. muscae species complex morphological and nuclear characteristics of the primary conidia are taxonomically important (Keller 1984). In this study we compared the dimensions of the primary conidia and their number of nuclei of isolates from their original (natural) hosts and after being transferred to alternative hosts (cross-transmission) in order to check the stability of these characteristics. We conclude that even though conidial characteristics change significantly when produced in alternative host species, the size and number of nuclei in the primary conidia are stable enough for species circumscriptions. Interestingly, the dimensions and number of nuclei in the primary conidia from isolates passed through Pollenia angustigena/P. rudis increased regardless of the original host species, but they were always within the range of the classically morphological species descriptions of either E. schizophorae or E. muscae s.str. (Keller et al. 1999). Host taxa have previously been shown to have an effect on the dimensions of the primary conidia as well as their number
of nuclei. However, the morphology of E. muscae isolates originating from different host taxa are not strongly divergent (Jensen et al. 2001), and conidial characteristics could not be used alone for diagnostic purposes due to the extent of their morphological variability. The shape of the hyphal bodies and yellowish colour of E. scatophagae conidia are the main phenetic criteria distinguishing E. scatophagae from E. muscae s.str. (Keller et al. 1999; Steinkraus & Kramer 1988). The yellowish colour of dry conidia, however, results from reflection of the yellow hair of the hirsute scatophagids (MacLeod et al.1976), and conidia of this species are uncoloured and hyaline on a glass slide. In addition, different post-mortem characteristicistics of flies have been suggested to have taxonomic value for separating Entomophthora species, but as these characteristics are host rather than fungus-dependent they must be considered to be invalid for taxonomic purposes (Steenberg et al. 2001). The phylogenetic sequence analyses of the E. muscae species complex resolves into three major lineages, each of which centres on a single species; E. muscae, E. schizophorae, and E. syrphi, that had been previously recognized using traditional phenetic and pathobiological characteristics (Ba1azy 1993; Keller 2002; Keller 1987). The E. syrphi and the E. muscae
Entomophthora muscae species complex
947
Table 5 – Transmission experiments of members of the Entomophthora muscae complex to different host taxa Fungus
Original host
Receptor fly
No. of cadavers
No. of repetitions
No. of flies at risk
No. of infected (%)
Coenosia tigrina (Muscidae) C. tigrina (Muscidae) C. tigrina (Muscidae)
C. tigrinac Delia radicum Musca domestica
2 8 12
2 8 12
9 15 53
0 (0.0) 2 (13.3) 30 (56.6)
M. M. M. M.
domestica (Muscidae) domestica (Muscidae) domestica (Muscidae) domestica (Muscidae)
M. domesticac C. tigrina D. radicum Pollenia sp.
39 5 72 9
8 1 35 5
138 19 85 28
133 (96.4) 0 (0.0) 3 (3.5) 5 (17.9)
E. muscae E. muscae
D. radicum (Anthomyiidae) D. radicum (Anthomyiidae)
D. radicumc M. domestica
34 66
18 50
87 272
16 (18.4) 222 (81.6)
E. muscae E. muscae E. muscae
Scatophaga stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae)
S. stercorariac D. radicum M. domestica
22 12 20
18 8 15
66 92 154
6 (9.1) 0 (0.0) 11 (7.1)
E. schizophoraeb E. schizophoraeb E. schizophoraeb
M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae)
M domestica c D. radicum Pollenia sp.
33 39 9
8 8 5
151 99 25
141 (93.4) 15 (15.2) 2 (8.0)
E. schizophorae E. schizophorae
Pollenia sp. (Calliphoridae) Pollenia sp. (Calliphoridae)
Pollenia sp.c M. domestica
12 18
12 17
68 93
44 (64.7) 51 (54.8)
E. syrphi E. syrphi E. syrphi
Indet. Syrphidae Indet. Syrphidae Indet. Syrphidae
Indet. Syrphidae D. radicum M. domestica
14 11 8
6 11 8
18 11 41
0 (0.0) 0 (0.0) 0 (0.0)
Entomophthora ferdinandi E. ferdinandi E. ferdinandi E. E. E. E.
muscae muscae muscae muscae
s.stra s.stra s.stra s.stra
All the cadavers used in the transmission experiments came from the field sampling, except for the Entomophthora muscae s.str. and E. schizophorae with M. domestica as original host. a The cadavers came from the in vivo culture DK3. b The cadavers came from the in vivo culture DPIL 150896. c The original host species.
s.str. lineages, which both include fungi having high numbers of nuclei in their conidia, were more closely related to each other than to the E. schizophorae lineage which has fewer nuclei in the conidia. This relationship should, however, be investigated further by including more Entomophthora species and also species affecting hosts other than dipterans. The molecular gave supported the existing taxonomy based on morphological characteristics and, together with ecological characteristics, provided evidence for a speciation that has not yet been recognized by morphological criteria. This is best exemplified in the E. muscae lineages, where sequence divergence of the ITS II and to a lesser extent the first part of the LSU rRNA gene was seen between the three isolates originating from three different host species. The multilocus approach RAPD that was used to investigate the genetic variation of members of the E. muscae lineages, particularly in E. scatophagae, also supported the cryptic speciation of E. muscae s.str. and agreed with the results from the UP-PCR, RAPD, and PCR-RFLP analyses by Jensen et al. (2001).
Operational taxonomic concept An operational species level concept for the taxonomic resolution of the Entomophthora muscae species complex that we now propose, seeks to accommodate non-taxonomists and researchers in applied science, as well as taxonomists. The genus Entomophthora is monophyletic evidenced by the ITS II and the LSU rRNA gene sequences analyses in the current
study and also by previous SSU rRNA gene sequences analyses (Freimoser et al. 2001; Jensen et al. 1998). However, the E. muscae species complex seems to be a paraphyletic assembly based on PCR-RFLP analysis (Jensen & Eilenberg 2001). Even though the host range for the multinucleate E. muscae group was expanded to also include Hymenoptera, we suggest that the E. muscae s.l. is restricted to species naturally found on derived clades of flies. E. muscae s.l. is, in this definition, not to be regarded as a single phylogenetic entity but as a workable assemblage. E. muscae s.str. from Musca domestica is genetically very different from isolates originating from other naturally infected dipteran host taxa and correspondingly significant phenetic differences were observed (Jensen et al. 2001). Therefore we suggest that the nomenclaturally strict sense of E. muscae should be limited to isolates originating from M. domestica, the host from which it was originally described (Cohn 1855) and redescribed (Keller et al. 1999). However, it would be interesting to study the biogeographic variation of E. muscae s.str. using geographically diverse samples of the globally distributed M. domestica. E. muscae isolates from hosts other than M. domestica, such as Delia radicum and Scatophaga stercoraria, clustered together in the molecular analyses on separate branches that do not include the fungi from M. domestica. Other studies have also shown that each host fly species had its own distinctive fungal pathogen genotype, but that they fall within the same morphological parameter ranges as E. muscae s.str. (Jensen et al. 2001; Thomsen & Jensen 2002). This indicates that a radial
948
A. B. Jensen et al.
92
E. muscae s. str (Md)
E. muscae s. str. (Md)
74
100
100 E. muscae (Dr) 74
92
E. scatophagae (Ss) 95
Entomophthora sp. (Mym)
E. schizophorae (Md) E. schizophorae (Pr)
E. muscae (Dr)
E. muscae (Dr)
97
E. scatophagae (Ss) Entomophthora sp. (Mym)
98
E. syrphi s. str. (Mem) 100
E. muscae s. str. (Md)
E. scatophagae (Ss) Entomophthora sp.
99
E. syrphi s. str. (Mem) E. schizophorae (Md)
100 50 changes
E. schizophorae (Pr)
E. syrphi s. str. (Mem) E. schizophorae (Md)
100 50 changes
E. schizophorae (Pr)
Entomophaga 50 changes
A
B
C
Fig 1 – Phylogenetic relationships within the Entomophthora muscae species complex inferred from parsimony analyses of the first part of the LSU rRNA gene (A), the ITS II (B) and a combined data set (C). BS percentages over 50 % from 1000 replicates are shown above each supported branch. A. The single most parsimonious phylogram requiring 226 steps (CI [ 0.9336 and RI [ 0.8235), Entomophaga aulicae was used as outgroup. B. The single most parsimonious phylogram requiring 500 steps (CI [ 0.9420 and RI [ 0.9254), midpoint rooting was used. C. The single most parsimonious phylogram requiring 726 steps (CI [ 0.9439 and RI [ 0.9297), midpoint rooting was used. Md [ Musca domestica; Dr [ Delia radicum; Ss [ Scatophaga stercoraria; Mym [ Myospila meditabunda; Mem [ Melanostoma mellinum; Pr [ Pollenia rudis.
adaptation toward high host specificity may have occurred, and that these ecological differences can be used in the taxonomic separation of E. muscae s.l. Based on the molecular data two possible taxonomic solutions exist: either E. muscae should be maintained as one highly plastic species (which would require rejection of E. scatophaga as a separate species), or E. muscae (as currently recognized within the E. muscae species complex) should be segregated into multiple new species representing the distinctive genotypes affecting each of the host fly species or genera. We recommend the latter solution. E. schizophorae is easily recognized by both morphological and molecular criteria. A certain overlap between the natural host range for E. schizophorae and E. muscae exists, e.g., in host taxa such as M. domestica and D. radicum, while in other host taxa, such as Chamaepsila rosae, only E. schizophorae infections have so far been found (Eilenberg 2002; Eilenberg & Philipsen 1988). The sequence differences between the M. domestica and P. rudis isolates used in this study suggest a similar divergence, as seen within the E. muscae lineage. But until a thorough study of the variation of E. schizophorae from different host taxa is performed E. schizophorae should be recognized as a single species. E. syrphi can be distinguished molecularly from the other members of the E. muscae species complex. Morphologically, however, it can be difficult to differentiate some E. syrphi from E. muscae s.str. (Steenberg et al. 2001), but E. syrphi generally has larger conidia containing more nuclei. E. syrphi seems to have a restricted host range and has only been recorded from a few hover fly species (Ba1azy 1993; Keller 1987). Variable RFLP patterns have been observed within E. syrphi, suggesting the existence of yet another unexplored species complex (Thomsen & Jensen 2002). We were not able to cross-infect E. syrphi from hover flies to other fly species. However, in a study by Steenberg et al. (2001) an Entomophthora with a high number
of conidial nuclei and recognized as E. syrphi from a muscoid fly (Phaonia perdita) was shown to cause infection in M. domestica. An Entomophthora species with similar morphology but isolated from another muscoid host species (Myospila meditabunda) was included in the current molecular analysis, where it formed a monophyletic group together with E. syrphi. Nevertheless, we do not recommend including these Entomophthora species in E. syrphi because of their high sequence divergence and their different host ranges. E. syrphi, as well as E. grandis, should be restricted to Entomophthora pathogenic to hoverflies. Moreover the other species deserves to receive an amplified and more detailed species description. A species can be described as a single lineage of ancestor– descendent populations which maintains its identity from
Scatophaga stercoraria
Delia radicum
Coenosia Musca domestica tigrina
P. infirma B. fugax Neg. control 1500 bp 1200 bp 1000 bp 800 bp 700 bp 500 bp 400 bp 300 bp
Fig 2 – Agarose gel electrophoresis of RAPD amplifications obtained with the primer OPA-10. A molecular size ladder appears in the first and last lane. The Entomophthora muscae isolates obtained from a single host species had similar profiles, whereas isolates from different host species had different profiles. P. [ Pegoplata, B. [ Botanophila.
Entomophthora muscae species complex
other such lineages and which has its own evolutionary tendencies and historical fate (Wiley 1978). For species within the E. muscae complex, genetic characteristics, physiological host range and morphology of the conidia are entities that can be used to develop an operational taxonomic species concept. This study shows that the various regions of the ribosomal repeat, in particular the ITS II, include variation suitable for distinguishing species within the E. muscae species complex. However, in future other genomic regions should be included to perpetuate the recognition of phylogenetically defined species (Taylor et al. 2000). Physiological host range as evidenced by transmission experiments only provides a hint of the ecological host range; however, the correlation of molecular data with phenotypic evidence and host range will continue to improve our understanding of the evolutionary processes and species limits within the Entomophthorales.
Acknowledgements We wish to thank Verner Michelsen and Stig Andersen, Zoological Museum, Copenhagen for identification of fly hosts. From the Royal Veterinary and Agricultural University, Jan Martin is thanked for help with collection of Scatophaga stercoraria, Kirsten Ploug and Mette Vingaard are thanked for technical assistance. Nicolai Vitt Meyling for comments on the manuscript and Charlotte Nielsen for statistical support. The Danish Pest and Infestation Laboratory is thanked for providing flies and two in vivo cultures. Richard Humber, USDA, Ithaca and Ingeborg Klingen, Norwegian Crop Research Institute are thanked for in vitro cultures. The Danish Ministry of Food, Agriculture and Fisheries, the Royal Veterinary and Agricultural University and the Carlsberg Foundation supported this work financially.
references
Baird RB, 1957. Notes on a laboratory infection of Diptera caused by the fungus Empusa muscae Cohn. Canadian Entomology 89: 432–435. Ba1azy S, 1993. Flora of Poland, Fungi (Mycota), Vol. 24, Entomophthorales. Wydawnictwa Inst. Bot. Pan., Krako´w. Beauvais A, Latge´ JP, 1988. A simple medium for growing entomophthoralean protoplasts. Journal of Invertebrate Pathology 51: 175–178. Ben-Ze’ev IS, Kenneth RG, 1982. Features-criteria of taxonomic importance in the Entomophthorales. Part I. A revision of the Batkoan classification. Mycotaxon 14: 393–455. Cohn F, 1855. Empusa muscae und die Krankheit der Stubenfliege. Hedwigia 10: 57–61. Collet D, 1991. Modeling Binary Data. Chapman & Hall, London. Eilenberg J, 2002. Biology of fungi from the order Entomophthorales with emphasis on the genera Entomophthora, Strongwellsea and Eryniopsis: A contribution to insect pathology and biological control. D.Sc. Thesis. Department of Ecology, The Royal Veterinary and Agricultural University, Copenhagen. Eilenberg J, Bresciani J, Martin J, 1987. Entomophthora species with E. muscae-like primary spores on two new insect orders, Coleoptera and Hymenoptera. Nordic Journal of Botany 7: 577–584.
949
Eilenberg J, Philipsen H, 1988. The occurrence of Entomophthorales on the carrot flies (Psila rosae F.) in the field during two successive seasons. Entomophaga 33: 135–144. Freimoser FM, Jensen AB, Tuor U, Aebi M, Eilenberg J, 2001. Isolation and in vitro cultivation of the aphid pathogenic fungus Entomophthora planchoniana. Canadian Journal of Microbiology 47: 1082–1087. Higgins DG, Bleasby AJ, Fuchs R, 1992. CLUSTAL V: improved software for multiple sequence alignment. Cabios 8: 89–191. Humber RA, 1981. An alternative view of certain taxonomic criteria used in the Entomophthorales (Zygomycetes). Mycotaxon 13: 191–240. Humber RA, 1989. Synopsis of a revised classification for the Entomophthorales (Zygomycotina). Mycotaxon 34: 441–460. Jensen AB, Eilenberg J, 2001. Genetic variation within the insectpathogenic genus Entomophthora, focusing on the E. muscae complex, using PCR-RFLP of the ITS II and LSU rDNA. Mycological Research 105: 307–312. Jensen AB, Gargas A, Eilenberg J, Rosendahl S, 1998. Relationships of the insect-pathogenic order Entomophthorales (Zygomycota, Fungi) based on phylogenetic analyses of nuclear small subunit ribosomal DNA sequences (SSU rDNA). Fungal Genetics and Biolology 24: 325–334. Jensen AB, Thomsen L, Eilenberg J, 2001. Intraspecific variation and host specificity of Entomophthora muscae s.str. isolates revealed by random amplified polymorphic DNA, universal primed PCR, PCR-restriction fragment length polymorphism, and conidial morphology. Journal of Invertebrate Pathology 78: 251–259. Kalsbeek V, Mullens BA, Jespersen JB, 2001. Field studies of Entomophthora (Zygomycetes: Entomophthorales) d induced behavioural fever in Musca domestica (Diptera: Muscidae) in Denmark. Biological Control 21: 264–273. Keller S, 1984. Entomophthora muscae as a species complex. Mittelung Schweizer Entomologische Gesellschaft 57: 131–132. Keller S, 1987. Arthropod-pathogenic Entomophthorales of Switzerland. I. Conidiobolus, Entomophaga and Entomophthora. Sydowia 40: 122–167. Keller S, 2002. The genus Entomophthora (Zygomycetes, Entomophthorales) with a description of five new species. Sydowia 54: 157–197. Keller S, Kalsbeek V, Eilenberg J, 1999. Redescription of Entomophthora muscae (Cohn) Fresenius. Sydowia 51: 197–209. Kjøller R, Rosendahl S, 2000. Detection of arbuscular mycorrhizal fungi (Glomales) in roots by nested PCR and SSCP (single stranded conformation polymorphism). Plant and Soil 226: 189–196. Klingen I, Meadow R, Eilenberg J, 2000. Prevalence of fungal infections in adult Delia radicum and Delia floralis captured on the edge of a cabbage field. Entomologia Experimentalis et Applicata 97: 265–274. Kramer JP, Steinkraus DC, 1981. Culture of Entomophthora muscae in vivo and its infectivity for six species of muscoid flies. Mycopathology 76: 139–143. MacLeod DM, Mu¨ller-Ko¨gler E, Wilding N, 1976. Entomophthora species with E. muscae-like conidia. Mycologia 68: 1–29. Lo´pez Lastra CC, Hajek AE, Humber RA, 2001. Effects of two cryopreservation techniques on viability and pathogenicity of entomophthoralean fungi. Canadian Journal of Botany 79: 861–864. Mullens BA, 1989. Cross-transmission of Entomophthora muscae (Zygomycetes: Entomophthoraceae) among naturally infected muscoid fly (Diptera: Muscidae) hosts. Journal of Invertebrate Pathology 53: 272–275. Mullens BA, Rodriguez JL, Meyer JA, 1987. An epizootiological study of Entomophthora muscae in muscoid fly populations on Southern California poultry facilities, with emphasis on Musca domestica. Hilgardia 55: 1–41.
950
Nagahama T, Sato H, Shimazu M, Sugiyama J, 1995. Phylogenetic divergence of the entomophthoralean fungi: evidence from nuclear 18 S ribosomal RNA gene sequences. Mycologia 87: 203–209. Ramoska WA, Hajek AE, Ramos ME, Soper RS, 1988. Infection of grasshoppers (Orthoptera, Acrididae) by members of the Entomophaga grylli species complex (Zygomycetes, Entomophthorales). Journal of Invertebrate Pathology 52: 309–313. Remaudie`re G, Keller S, 1980. Re´vision syste´matique des genres d’Entomophthoraceae a` potentialite´ entomopathoge`ne. Mycotaxon 11: 323–338. SAS Institute Inc, 1999. SAS/STAT User’s Guide, 1st edn., Vol. 2, SAS Institute Inc., Cary, NC, USA, Version 8. Steenberg T, Jespersen JB, Jensen KMV, Nielsen BO, Humber RA, 2001. Entomopathogenic fungi in flies associated with pastured cattle in Denmark. Journal of Invertebrate Pathology 77: 186–197. Steinkraus DC, Geden CJ, Rutz DA, 1993. Prevalence of Entomophthora muscae (Cohn) Fresenius (Zygomycetes: Entomophthoraceae) in houseflies (Diptera: Muscidae) on dairy farms in New York and induction of epizootics. Biological Control 3: 93–100. Steinkraus DC, Kramer JP, 1987. Susceptibility of sixteen species of Diptera to the fungal pathogen Entomophthora muscae (Zygomycetes: Entomophthoraceae). Mycopathology 100: 55–63. Steinkraus DC, Kramer JP, 1988. Entomophthora scatophagae desc. ampl. (Zygomycetes: Entomophthorales), a fungal pathogen of the
A. B. Jensen et al.
yellow dung fly, Scatophaga stercoraria (Diptera: Anthomyiidae). Mycotaxon 32: 105–113. Swofford DL, 1998. PAUP: Phylogenetic Analysis Using Parsimony [Version 4]. Sinauer Assosiates, Sunderland, MA. Taylor JW, Jacobson DJ, Kroken S, Kasuga T, Geiser DM, Hibbett DS, Fisher MC, 2000. Phylogenetic species recognition and species concepts in fungi. Fungal Genetic and Biology 31: 21–32. Thomsen L, Jensen AB, 2002. Application of nested-PCR technique on resting spores from the Entomophthora muscae species complex: implications for analyses of host–pathogen population interactions. Mycologia 94: 794–802. Van de Peer Y, De Wachter R, 1994. TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Computer Applications in the Biosciences 10: 569–570. Watson DW, Petersen JJ, 1993. Seasonal activity of Entomophthora muscae (Zygomycetes: Entomophthorales) in Musca domestica L. (Diptera: Muscidae) with reference to temperature and relative humidity. Biological Control 3: 182–190. Wiley EO, 1978. The evolutionary species concept reconsidered. Systematic Zoology 27: 17–26. White TJ, Bruns TD, Lee SB, Taylor JW, 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JS, White TJ (eds), PCR Protocols: A Guide To Methods And Applications. Academic Press, San Diego, pp. 315–321.
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/mycres
Value of host range, morphological, and genetic characteristics within the Entomophthora muscae species complex Annette Bruun JENSEN*, Lene THOMSEN, Jørgen EILENBERG Department of Ecology, The Royal Veterinary and Agricultural University (KVL), Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark
article info
abstract
Article history:
Entomopthora muscae sensu lato is a complex of morphologically similar fungal species path-
Received 25 November 2005
ogenic to evolutionarily advanced flies (Cyclorrhapha). To reach an operational species
Received in revised form
definition and recognition of species within this complex, the values of host range,
25 April 2006
morphological and genetic characteristics are reconsidered. Within the E. muscae species
Accepted 1 June 2006
complex morphological and nuclear characteristics of the primary conidia are taxonomi-
Published online 14 August 2006
cally important. In this study we compared the dimensions and nuclear numbers of the
Corresponding Editor:
primary conidia of isolates from their original (natural) hosts and after being transferred
Richard A. Humber
to alternative hosts (cross-transmission) in order to check the stability of these characteristics. The conidial characteristics change substantially when produced in alternative host
Keywords:
species, but their overall range in variability still fit within the traditional morphological
Anthomyiidae
species circumscriptions. The phylogenetic analyses of the ITS II and LSU rRNA gene
Diptera
sequences, revealed three distinct lineages within the complex: E. schizophorae, E. muscae
Ecology
and E. syrphi. Within each of these lineages sequence divergence was seen between isolates
Entomophthorales
originating from different host species. Our studies on the physiological host range showed
Host range
that several isolates were able to infect alternative dipteran species. Musca domestica was
Insect pathogenic fungi
a particularly good receptor. The ecological host range of any individual isolate seems,
Morphology
however, to be limited to one host species evidenced by the occurrence of distinct geno-
Muscidae
types within each natural infected host species shown by RAPD. The high host specificity
Phylogeny
of these fungi emphasizes the importance of identifying the host taxon at species level in
Taxonomy
the recognition of Entomophthora species. We recommend that morphological characteris-
Zygomycota
tics of fungal structures and host taxon, together with molecular data, serve as criteria for species determination in future studies on members of the E. muscae complex. ª 2006 The British Mycological Society. Published by Elsevier Ltd. All rights reserved.
Introduction Fungi from the Entomophthora muscae species complex are important for the natural population regulation of many dipteran species. Epizootics have often been reported also among pest flies, and the fungi possess a potential to be used in biocontrol
(Klingen et al. 2000; Mullens et al. 1987; Steinkraus et al. 1993; Watson & Petersen 1993). However, the evaluation of this potential requires a more complete understanding of the ecological interaction between host species and their pathogens. In such ecological studies the circumscriptions of individual fungal taxa become a central issue. Traditionally, species were
* Corresponding author. Tel.: þ45 35 28 26 82; fax: þ45 35 28 26 70 E-mail address: [email protected]. 0953-7562/$ – see front matter ª 2006 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.mycres.2006.06.003
942
described based on recognizable morphological features and as such E. muscae from Musca domestica was described by Cohn in 1855. Subsequently, fungal specimens from flies of derived clades (Cyclorrhapha) with E. muscae-like conidia were designated as E. muscae (MacLeod et al. 1976). In the 1980s cytological features were considered important for the taxonomy of the order Entomophthorales (Ben-Ze’ev & Kenneth 1982; Humber 1981, 1989; Remaudie`re & Keller 1980), in particular the contents of heterochromatin, number and size of the nuclei. Inspired by these new criteria, Keller (1984) suggested that E. muscae s.l. is a complex of species separated by conidial size, number and size of the nuclei. At present six species are recognized and described morphologically: E. ferdinandi, E. grandis, E. muscae s.str., E. scatophagae, E. schizophorae, and E. syrphi (Keller 2002). Some of the species, e.g., E. schizophorae, E. ferdinandi, and E. grandis, have distinct conidial morphology whereas others such as E. muscae s.str. and E. scatophagae are virtually indistinguishable using conidial characteristicistics. The two latter species can only be differentiated morphologically by the shape of their hyphal bodies, which are reported to be consistently spherical in E. muscae s.str. (Keller et al.1999). Whether this morphological variation is stable or due to phenotypic modifications associated with the environmental condition (the host species) remains unknown. Host range has been used to elucidate entomophthoralean species complexes. Based on transmission experiments, RFLP, and hybridization of a DNA probe, Entomophaga grylli has been divided into pathotypes according to host subfamily (Ramoska et al. 1988). Several laboratory transmission experiments of E. muscae s.l. have been performed within a range of Diptera (Baird 1957; Eilenberg et al. 1987; Kramer & Steinkraus 1981; Mullens 1989; Steenberg et al. 2001; Steinkraus & Kramer 1987), but no attempt to group E. muscae into pathotypes or groups based on their host range has been made. However, E. scatophagae is believed to have a restricted host range and is only described from a single host species; E. grandis and E. syrphi have only been reported from larger and smaller syrphid flies, respectively (Keller 2002). A recent study of the intraspecific variation within E. muscae s.str. showed that several genotypes occur and that each type was restricted to a single host species (Jensen et al. 2001). This indicates a high degree of host specificity in the E. muscae species complex similar to that observed in the E. grylli complex. Phylogenetic relationships of the order Entomophthorales have been inferred from the SSU rRNA gene sequences (Jensen et al. 1998; Nagahama et al. 1995); but until now no sequence analyses have been accomplished to resolve species complexes within Entomophthorales. Nevertheless, PCR-RFLP of the ITS II and the 50 end of the LSU rRNA gene support the separate origin of E. muscae, E. schizophorae, and E. syrphi (Jensen & Eilenberg 2001). The aim of this study was to evaluate the value of host range, morphological, and genetic characteristics for operational species recognition of species within the E. muscae complex. Studies were made to elucidate specific topics: (1) the plasticity of certain morphological characteristics was explored, (2) the physiological host range among certain host taxa was examined, and (3) the phylogeny was assessed by sequence analyses of the ITS II and part of the LSU rRNA gene.
A. B. Jensen et al.
Materials and methods Field samplings Approximately 8000 live flies (Table 1) were collected by sweep-net in Denmark during 1999 (June–September) and 2000 (May–October) from the following habitats: cabbage and rape fields, border vegetation, forests, stables, and houses. Live flies were transferred individually to 30 ml cups containing 5 ml 1.5 % water-agar as described by Eilenberg & Philipsen (1988). Condensation in the cup was avoided by cutting a hole in the lid and covering it with gauze. Food (a paste of 9 parts sugar, 1 part dry yeast, and little water) was applied directly onto the gauze. Predatory flies (predominantly Coenosia tigrina and Scatophaga stercoraria) were also supplied with live houseflies every third day, and hover flies (Syrphidae) were provided with flowers or pollen. The flies were incubated at room temperature and exposed to natural day length. They were examined for death and external signs of fungal infection every day in a 14 d period. Conidia were collected on glass slides in a humid chamber from all the cadavers and stored dry until morphological examination. Immediately after the conidial collection, well-sporulating cadavers were either used in the host range transmission experiments or were used to obtain in vitro isolates to ensure maximum inoculum levels.
Morphological characteristics The morphology of primary conidia discharged from the original hosts was compared with the morphology of the same in vivo-isolate after cross-transmission to an alternative host. The dimensions of 20 primary conidia per isolate were measured mounted in lactophenol-cotton blue (0.01 % cotton blue) using an Olympus Provis microscope supplemented with an Oly-Lite computer-based system for morphometrics at 400 magnification. The number of nuclei of 20 primary conidia was counted mounted in DAPI (4,6-diamidino-2-phenylindole, 5 mg ml1) using a Zeiss Axiophot epifluorescence microscope (filter 0.5; violet, excitation 395–440) at 400 magnification.
Table 1 – Number of collected flies and number of flies with external sign of Entomophthora muscae s.l. Host taxa
Anthomyiidae Anthomyiidae sp. Delia radicum Calliphoridae Pollenia angustigena/rudis Muscidae Coenosia tigrina Musca domestica Scatophagidae Scatophaga stercoraria Syrphidae Syrphidae sp.1
Approx. number of collected flies
Number of Entomophthora infected flies
350 1100
25 250
900
90
300 700
20 100
1650
90
1000
90
1 From the genera Melanostoma and Platycheirus.
Entomophthora muscae species complex
943
Standard LSD tests (PROC GLM; SAS Institute 1999) were used to test if the interaction of host and isolate had significant influence on the length, width, and number of nuclei in the primary conidia of Entomophthora muscae and E. schizophorae. Within each isolate comparison of mean values were made using the Tukey test.
this study were stored in the Fungal Insect Pathogenic Culture Collection at the Royal Veterinary and Agricultural University, Copenhagen, by cryopreservation at 80 C according to Lo´pez Lastra et al. (2001) and are deposited in ARSEF (USDAARS, Ithaca, NY). All the isolates used are listed in Tables 2 and 3.
In vitro isolation
Physiological host range
In vitro isolation of the Entomophthora species was done as in Jensen et al. (2001). Fresh, well-sporulating cadavers were placed on a small sterile inverted Petri dish lid (3.5 cm diam) and the base was put on top. The wings of the large flies, Musca domestica, Myospila meditabunda, Pollenia angustigena, P. rudis, and Scatophaga stercoraria were cut off in order to avoid contamination by wings hitting the Petri dish base. Sporulation was allowed for 30 min and the arrangement was placed in a humid chamber in order to enhance sporulation. In a sterile hood 1 ml GLEN (Beauvais & Latge´, 1988) including 5 % fetal bovine serum was poured onto the discharged conidia. A new sterile lid was used and the petri dish was sealed with Parafilm and incubated dark at 23 C. When growth was observed, the in vitro culture was transferred to 50 ml cell culture flasks with 5 ml GLEN for further growth. The isolates used in
Laboratory transmission experiments were conducted to obtain information on physiological host range. Two types of transmission experiments were performed: in small cup experiments, one to five live flies (receptors) and a wellsporulating cadaver (donor) were placed together in a 30 ml cup containing 5 ml 1.5 % water-agar. Cups with live flies and sporulating cadavers were maintained at 18 C and received a photoperiod of 16:8 (light:dark) in 24 h for fungal inoculation. Alternatively, in caged experiments, ten to 20 live flies (receptors) were placed in a cylindrical cardboard container (9 cm diam, 5 cm high). The cardboard containers had a Petri dish base (9 cm diam) as the bottom and a 2-mm mesh top as a cover. Two to five well-sporulating cadavers (donors) were placed on the mesh top and a Petri dish lid (9 cm diam) was put on top of the sporulating cadavers. The
Table 2 – Insect host and geographic origin of the in vitro isolates used for RAPD Isolatea KVL 00-56 KVL 00-57b KVL 00-58 KVL 00-59 KVL 00-60 (ARSEF 6704) KVL 00-61 (ARSEF 6705) KVL 00-62 KVL 00-63 KVL 00-64 KVL 00-65 KVL 00-66 KVL 00-67 NCRI 1-99 (ARSEF 6274) ARSEF 2668 KVL 99-102 (ARSEF 6815)b KVL 99-103 KVL 99-96 KVL 99-85(ARSEF 6811) KVL 99-13 (ARSEF 6132)bc KVL 99-23 (ARSEF 6141) KVL 99-110 (ARSEF 6813) ARSEF 2542 KVL 99-63 KVL 99-70 KVL 99-90 KVL 99-92 (ARSEF 6918) KVL 99-87 NCRI 2-99
Fungal species
Host species (host family)
Collection site
Year
Entomophthora scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae E. scatophagae Entomophthora muscae E. muscae E. muscae E. muscae E. muscae E. muscae E. muscae s.str. E. muscae s.str. E. muscae s.str. E. muscae s.str. E. muscae s.str. E. muscae s.str. E. ferdinandi E. ferdinandi E. ferdinandi E. ferdinandi
Scatophaga stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) Delia radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) D. radicum (Anthomyiidae) Musca domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae) Coenosia tigrina (Muscidae) C. tigrina (Muscidae) Pegoplata infirma (Anthomyiidae) Botanophila fugax (Anthomyiidae)
Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Tuse Næs, Denmark Ski, Norway Svanholm, Denmark Sengeløse, Denmark Ørslev, Denmark Sengeløse, Denmark Hegnstrup, Denmark Kirke Sa˚by, Denmark Skævinge, Denmark Knardrup, Denmark Ribe, Denmark Sverkildstup, Denmark Sverkildstup, Denmark Ørslev, Denmark Ørslev, Denmark Hegnstrup, Denmark Ski, Norway
2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 1999 1986 1999 1999 1999 1999 1999 1999 1999 1986 1999 1999 1999 1999 1999 1999
a KVL, Danish abbreviation for the Royal Veterinary and Agricultural University, Copenhagen; ARSEF, ARS Collection of Entomopathogenic Fungi, Ithaca; NCRI, Norwegian Crop Research Institute. b Isolates also used for the sequence analyses. c In vitro isolate made from in vivo cultures, that was also used for the transmission experiments.
KVL, Danish abbreviation for the Royal Veterinary and Agricultural University, Copenhagen; ARSEF, ARS Collection of Entomopathogenic Fungi, Ithaca. Isolates also used for the RAPD analysis. In vitro isolate made from in vivo cultures, that was also used for the transmission experiments. Genbank accession numbers for the sequences. a b c d
1999 1999 2000 1996 2000 2000 1998 Kirke Sa˚by, Denmark Sengeløse, Denmark Tuse Næs, Denmark Roskilde, Denmark Hundested, Denmark Hundested, Denmark Gribskov Denmark Musca domestica (Muscidae) Delia radicum (Anthomyiidae) Scatophaga stercoraria (Scatophagidae) M. domestica (Muscidae) Pollenia rudis (Calliphoridae) M. meditabunda (Muscidae) Melanostoma mellinum (Syrphidae) KVL KVL KVL KVL KVL KVL KVL
99-13 (ARSEF 6132)b,c 99-102 (ARSEF 6815)b 00-57 (ARSEF 6704)b 99-18 (ARSEF 6137) 00-93 (ARSEF 6817) 00-92 (ARSEF 6701) 98-19 (ARSEF 5955)
Entomophthora muscae s.str. E. muscae E. scatophagae E. schizophorae E. schizophorae Entomophthora sp. E. syrphi s.str.
16–18 14–16 16–18 5–7 5–7 20–23 17–21
DQ481217/DQ481224 DQ481218/DQ481225 DQ481219/DQ481226 DQ481220/DQ481227 DQ481221/DQ481228 DQ481222/DQ481229 DQ481223/DQ481230
Year Host (family) Isolatea
Species
Nuclei number
Accession no.d ITSII/LSU
Collection site
A. B. Jensen et al.
Table 3 – Insect host and geographic origin of the isolates used for sequencing of the ITS II and the first part of the LSU rRNA gene
944
cardboard assemblies with live flies and sporulating cadavers were maintained 18 C and received a photoperiod of 16:8 (light:dark) in 24 h for fungal inoculation. The cage type was only performed with donors from two in vivo cultures of Entomophthora muscae s. str (DK3) (Keller et al. 1999) and E. schizophorae (DPIL 150896) (Kalsbeek et al. 2001) and with Musca domestica and Delia radicum as receptors. The two in vivo cultures were maintained by continuous passages in M. domestica as described by Kramer and Steinkraus (1981), and both had M. domestica as original host. After the inoculation, the flies were placed individually in 30 ml cups containing water agar and covered with gauze and a lid with a hole. The sugar–yeast paste described above was provided as food directly on the gauze. The flies in the cups were incubated at 18 C with a photoperiod of 16:8 (light:dark) for three weeks, and were examined daily for death and external signs of fungal infection. Conidia from the cadavers were collected and stored dry until examination. M. domestica originated from a laboratory stock and was supplied as pupae by the Danish Institute of Agricultural Sciences, Department of Integrated Pest Management, Danish Pest Infestation Laboratory Group (DPIL). The other receptor flies originated from the field but were used only after a quarantine period of at least two weeks to exclude contamination from natural infections. Thus the field collected flies were old and therefore those that died during the transmission experiments without external signs of Entomophthora ssp. infections were excluded. Logistic regression was used to analyse the effects of fungal species and the receptor fly on the percentage of infected flies (Proc GENMOD; SAS Institute 1999) with a binomial distribution and logit as link functions. Over-dispersion was taken into account (Collet 1991). To test the effect of receptor fly, all receptor flies were categorized into three groups: (1) the receptor fly was of the same species as the donor fly, (2) the receptor fly was M. domestica, and (3) the receptor fly was neither the same fly species as the donor fly or M. domistica.
Genetic characteristics DNA extraction and amplification of ITS II and the LSU rRNA gene of seven in vitro isolates (Table 2) were conducted as described by Jensen et al. (2001). For the amplification of the ITS II the primers nu-5.8 S-50 (Jensen & Eilenberg 2001) and ITS 4 (White et al. 1990) were used and for the amplification of the LSU the primers nu-LSU 0018-50 (Jensen & Eilenberg 2001) and nu-LSU 0805-30 (Kjøller & Rosendahl 2000) were used. The PCR conditions were initial denaturation for 5 min at 96 C, followed by 35 cycles with denaturation for 1 min at 96 C, annealing for 1 min at 62 C (ITS II) or 55 C (LSU), extension for 1 min at 72 C and a final extension for 10 min at 71 C. The PCR reactions were carried out in a 25 ml volume with 250 mM of each dNTP, 0.8 mM of each primer, 2.5 mM MgCl2, 1 buffer (10 mM Tris–HCl, pH 8.8 at 25 C, 50 mM KCl, 0.1 % Triton X-100), 1 unit DyNazyme II (Finnzymes, Espoo, Finland) and 1 ml extracted DNA, diluted at 1:10 or 1:100. Before sequencing, the PCR products were purified with the GFXtm PCR DNA and Gel Purification Kit (Amersham Bioscience, Uppsala, Sweden). Both strands of the ITS II and LSU were sequenced by GATC GmbH. To get the full sequence of the ITS II
Entomophthora muscae species complex
four internal primers ITS II 500-50 (5’-ACGTTGGCAACGAC WAGTAA-3’), ITS II 500-3’ (5’-TTACTWGTCGTTGCCAACGT3’), ITS II 1000-5’ (5’-TTGTTGAAWCTTTTCTGTCGC-3’) and ITS II 1000-30 (50 -CGACAGAAAAGWTTCAACAAT-30 ) were used. The chromatograms were checked using Sequencer 3.1 (Gene Codes Corporations, Ann Arbor, MI). In addition the LSU sequence of Entomophaga aulicae (Genbank accession no: U35394) was included, but the E. aulicae ITS II sequence was too divergent to be aligned with the sequences generated in this study. Following initial alignment with CLUSTAL V 1.60 (Higgins et al. 1992) the alignments were adjusted manually. MP analyses were performed with the PAUP 4.0b6 (Swofford 1998) on individual data sets (ITS II and LSU) and combined data set (ITS II þ LSU). The branch-and-bound search option was used to find exact solutions. All characteristics were equally weighted, and invariant characteristics were ignored. In the LSU analysis E. aulicae was chosen as an outgroup. Supports for internal branches were assessed by 1000 BS replications, and the partition-homogeneity test was used to evaluate the concordance between two individual data sets. NJ analyses were performed with the software Treecon for Windows (Van de Peer & De Wachter 1994) using the JukesCantor evolutionary model. Supports for internal branches were assessed by 1000 BS replications. RAPD was performed with the 10 commercial RAPD primers (OPA01, OPA02, OPA05, OPA06, OPA07, OPA09, OPA10, OPA15, OPA17,OPA20; Operon Technologies, Alameda, CA) on 11 Entomophthora scatophagae, six E. muscae, six E. muscae s.str., and four E. ferdinandi in vitro isolates originating from various host species (Table 2) as described in Jensen et al. (2001).
Results Morphological characteristics The primary conidia all were campanulate and multinucleate irrespective of host origin. The overall analysis of the primary conidia from the cross-transmission experiments showed significant effects on the interaction of host species and isolate on length (F4, 266 ¼ 75.80; P < 0.0001), diameter (F4, 266 ¼ 49.16; P < 0.0001), and number of nuclei (F4, 266 ¼ 3.83; P ¼ 0.0048) for Entomophthora muscae and on length (F1, 133 ¼ 19.19; P < 0.0001) and diameter (F1, 133 ¼ 89.78; P < 0.0001) for E. schizophorae. The number of nuclei in E. schizophorae varied a little, but was not significantly different irrespective of host species (F1, 133 ¼ 1.29; P ¼ 0.2587). No clear pattern in size or number of nuclei after transmission to alternative hosts was detected, except for Pollenia angustigena/P. rudis where conidia from the original host were slightly larger and contained more nuclei than conidia of the same isolates produced from other dipteran hosts (Table 4). However, the changes in morphology after transmission to alternative hosts were never so dramatic that they influenced the classical morphological species identification.
Physiological host range Laboratory transmission of Entomophthora muscae s.l. between different host species was possible, but with variable success
945
(Table 5). The mortality attributable to the entomophthoralean fungi was significantly affected by the receptor fly (c2 ¼ 15.89; df ¼ 2; P ¼ 0.0004). Having Musca domestica as receptor fly compared with having the same receptor as donor fly species did not show significantly increased infectivity, as tested using Pair-wise comparisons using least squares means to test the effects of the species of receptor fly [Pr > |t| for H0: LSMean (category i) ¼ LSMean (category j)] (c2 ¼ 0.55; df ¼ 1; P ¼ 0.4593), whereas all other comparisons were significantly different (P 0.0089). Whereas M. domestica was an excellent receptor for all isolates tested (except for those from E. syrphi), Coenosia tigrina proved difficult to infect in our study conditions, even with E. muscae from naturally infected C. tigrina. Interestingly we observed a decrease in the number of conidiophores, as only one or two narrow abdominal bands of conidiophores appeared in several of the alternative hosts, thus resulting in moderate to restricted sporulation.
Genetic characteristics Approximately 1440 bp of the ITS II and 900 bp of the 50 end of the LSU rRNA gene were sequenced from the seven isolates listed in Table 3. The ITS alignment had 306 informative sites, whereas the LSU alignment, including Entomophthora aulicae, had 56 informative sites. Both regions produced trees with similar topology using maximum parsimony (Fig 1A–B). The parsimony-based partition homogeneity test found no support for incongruence between these two data sets (P ¼ 1.000), and thus the two data sets were combined. The tree topology of the combined data sets was similar to the tree topology of the individual data sets (Fig 1C), as were also the topology of the NJ trees (data not shown). Three major lineages of the E. muscae species complex were resolved, each of which centres on a single species; E. muscae, E. schizophorae, and E. syrphi. Each of the lineages was supported by BS values above 92 %. The E. syrphi and E. muscae lineages clustered together on a branch supported by a BS value of 74 % in the LSU analysis. In this study E. scatophagae was isolated in vitro for the first time. All 11 E. scatophagae isolates obtained had similar RAPD profiles with the ten RAPD primers used, but were different from E. muscae isolates originating from other host species (Fig 2).
Discussion Host range has long been regarded to be an important characteristic in the circumscription of Entomophthora species. Physiological host range can be reflected in laboratory transmission experiments, and our study showed that receptor flies from other (unnatural) host taxa generally were less susceptible than the original (natural) host flies, which is concordant to other transmission experiments performed (Baird 1957; Eilenberg et al. 1987; Kramer & Steinkraus 1981; Mullens 1989; Steenberg et al. 2001; Steinkraus & Kramer 1987). Those specimens that got infected often showed a decreased sporulation, which would consequently decrease further disease transmission. However, Musca domestica was equally susceptible to E. muscae s.l., compared with the exposed original host species.
946
A. B. Jensen et al.
Table 4 – Dimensions and numbers of nuclei in primary conidia of Entomophthora muscae and E. schizophorae before and after transmission to alternative host species Transmission experiments
Fungal species
Host species
Mean no. of nuclei
Mean length (mm)
Mean width (mm)
1 1
Entomophthora muscae E. muscae
Scatophaga stercorariaa Musca domestica
17.4 a 15.6 b
27.2 a 28.2 b
21.8 a 22.4 a
2 2
E. muscae E. muscae
Delia radicuma M. domestica
17.1 a 16.1 b
29.2 b 25.8 a
22.6 b 20.1 a
3 3 3
E. muscae E. muscae E. muscae
D. radicuma M. domestica M. domestica / D. radicumb
16.6 a 15.6 b 15.6 b
24.6 a 27.1 c 25.6 b
18.5 a 22.3 c 20.5 b
4 4
E. muscae E. muscae
D. radicuma M. domestica
14.8 a 15.5 a
23.7 a 25.7 b
18.5 a 19.4 b
5 5
E. muscae E. muscae
D. radicuma M. domestica
15.3 a 16.0 a
24.3 a 28.2 b
18.5 a 23.2 b
6 6 6
E. muscae s.str.d E. muscae s.str.d E. muscae s.str.d
M. domesticaa D. radicum Pollenia angustigena/rudisc
16.7 b 17.0 b 19.4 a
29.7 b 25.4 a 31.4 c
22.8 b 20.7 a 25.1 c
7 7 7
E. schizophoraee E. schizophoraee E. schizophoraee
M. domesticaa D. radicum P. angustigena/rudisc
5.5 a 5.5 a 5.6 a
22.6 a 23.2 a 23.9 b
17.9 a 18.0 ab 18.5 b
8 8
E. schizophorae E. schizophorae
Coenosia tigrinaa M. domestica
5.6 a 5.5 a
20.4 a 20.5 a
14.5 a 15.1 b
9 9
E. schizophorae E. schizophorae
P. angustigenaa M. domestica
6.1 a 5.5 a
23.8 a 20.7 b
19.3 b 15.2 a
All the cadavers used in the transmission experiments came from the field sampling, except the for the Entomophthora muscae s.str. and E. schizophorae with M. domestica as original host. Within each transmission experiment comparison of the mean values were made using a Tukey test and mean values followed by different letters are significantly different (P < 0.05). a The original host species. b This isolate was transmitted from Delia radicum to Musca domestica and then back to D. radicum again. c A mixed population of Pollenia angustigena and P. rudis. d The cadavers came from the in vivo culture DK3. e The cadavers came from the in vivo culture DPIL 150896.
Nevertheless, only a single genotype of E. muscae s.str. was previously documented from Danish collections of M. domestica (Jensen et al. 2001) sampled at different locations and years, so the high susceptibility of M. domestica to several E. muscae genotypes seems not to be reflected in natural epizootics. Within the E. muscae species complex morphological and nuclear characteristics of the primary conidia are taxonomically important (Keller 1984). In this study we compared the dimensions of the primary conidia and their number of nuclei of isolates from their original (natural) hosts and after being transferred to alternative hosts (cross-transmission) in order to check the stability of these characteristics. We conclude that even though conidial characteristics change significantly when produced in alternative host species, the size and number of nuclei in the primary conidia are stable enough for species circumscriptions. Interestingly, the dimensions and number of nuclei in the primary conidia from isolates passed through Pollenia angustigena/P. rudis increased regardless of the original host species, but they were always within the range of the classically morphological species descriptions of either E. schizophorae or E. muscae s.str. (Keller et al. 1999). Host taxa have previously been shown to have an effect on the dimensions of the primary conidia as well as their number
of nuclei. However, the morphology of E. muscae isolates originating from different host taxa are not strongly divergent (Jensen et al. 2001), and conidial characteristics could not be used alone for diagnostic purposes due to the extent of their morphological variability. The shape of the hyphal bodies and yellowish colour of E. scatophagae conidia are the main phenetic criteria distinguishing E. scatophagae from E. muscae s.str. (Keller et al. 1999; Steinkraus & Kramer 1988). The yellowish colour of dry conidia, however, results from reflection of the yellow hair of the hirsute scatophagids (MacLeod et al.1976), and conidia of this species are uncoloured and hyaline on a glass slide. In addition, different post-mortem characteristicistics of flies have been suggested to have taxonomic value for separating Entomophthora species, but as these characteristics are host rather than fungus-dependent they must be considered to be invalid for taxonomic purposes (Steenberg et al. 2001). The phylogenetic sequence analyses of the E. muscae species complex resolves into three major lineages, each of which centres on a single species; E. muscae, E. schizophorae, and E. syrphi, that had been previously recognized using traditional phenetic and pathobiological characteristics (Ba1azy 1993; Keller 2002; Keller 1987). The E. syrphi and the E. muscae
Entomophthora muscae species complex
947
Table 5 – Transmission experiments of members of the Entomophthora muscae complex to different host taxa Fungus
Original host
Receptor fly
No. of cadavers
No. of repetitions
No. of flies at risk
No. of infected (%)
Coenosia tigrina (Muscidae) C. tigrina (Muscidae) C. tigrina (Muscidae)
C. tigrinac Delia radicum Musca domestica
2 8 12
2 8 12
9 15 53
0 (0.0) 2 (13.3) 30 (56.6)
M. M. M. M.
domestica (Muscidae) domestica (Muscidae) domestica (Muscidae) domestica (Muscidae)
M. domesticac C. tigrina D. radicum Pollenia sp.
39 5 72 9
8 1 35 5
138 19 85 28
133 (96.4) 0 (0.0) 3 (3.5) 5 (17.9)
E. muscae E. muscae
D. radicum (Anthomyiidae) D. radicum (Anthomyiidae)
D. radicumc M. domestica
34 66
18 50
87 272
16 (18.4) 222 (81.6)
E. muscae E. muscae E. muscae
Scatophaga stercoraria (Scatophagidae) S. stercoraria (Scatophagidae) S. stercoraria (Scatophagidae)
S. stercorariac D. radicum M. domestica
22 12 20
18 8 15
66 92 154
6 (9.1) 0 (0.0) 11 (7.1)
E. schizophoraeb E. schizophoraeb E. schizophoraeb
M. domestica (Muscidae) M. domestica (Muscidae) M. domestica (Muscidae)
M domestica c D. radicum Pollenia sp.
33 39 9
8 8 5
151 99 25
141 (93.4) 15 (15.2) 2 (8.0)
E. schizophorae E. schizophorae
Pollenia sp. (Calliphoridae) Pollenia sp. (Calliphoridae)
Pollenia sp.c M. domestica
12 18
12 17
68 93
44 (64.7) 51 (54.8)
E. syrphi E. syrphi E. syrphi
Indet. Syrphidae Indet. Syrphidae Indet. Syrphidae
Indet. Syrphidae D. radicum M. domestica
14 11 8
6 11 8
18 11 41
0 (0.0) 0 (0.0) 0 (0.0)
Entomophthora ferdinandi E. ferdinandi E. ferdinandi E. E. E. E.
muscae muscae muscae muscae
s.stra s.stra s.stra s.stra
All the cadavers used in the transmission experiments came from the field sampling, except for the Entomophthora muscae s.str. and E. schizophorae with M. domestica as original host. a The cadavers came from the in vivo culture DK3. b The cadavers came from the in vivo culture DPIL 150896. c The original host species.
s.str. lineages, which both include fungi having high numbers of nuclei in their conidia, were more closely related to each other than to the E. schizophorae lineage which has fewer nuclei in the conidia. This relationship should, however, be investigated further by including more Entomophthora species and also species affecting hosts other than dipterans. The molecular gave supported the existing taxonomy based on morphological characteristics and, together with ecological characteristics, provided evidence for a speciation that has not yet been recognized by morphological criteria. This is best exemplified in the E. muscae lineages, where sequence divergence of the ITS II and to a lesser extent the first part of the LSU rRNA gene was seen between the three isolates originating from three different host species. The multilocus approach RAPD that was used to investigate the genetic variation of members of the E. muscae lineages, particularly in E. scatophagae, also supported the cryptic speciation of E. muscae s.str. and agreed with the results from the UP-PCR, RAPD, and PCR-RFLP analyses by Jensen et al. (2001).
Operational taxonomic concept An operational species level concept for the taxonomic resolution of the Entomophthora muscae species complex that we now propose, seeks to accommodate non-taxonomists and researchers in applied science, as well as taxonomists. The genus Entomophthora is monophyletic evidenced by the ITS II and the LSU rRNA gene sequences analyses in the current
study and also by previous SSU rRNA gene sequences analyses (Freimoser et al. 2001; Jensen et al. 1998). However, the E. muscae species complex seems to be a paraphyletic assembly based on PCR-RFLP analysis (Jensen & Eilenberg 2001). Even though the host range for the multinucleate E. muscae group was expanded to also include Hymenoptera, we suggest that the E. muscae s.l. is restricted to species naturally found on derived clades of flies. E. muscae s.l. is, in this definition, not to be regarded as a single phylogenetic entity but as a workable assemblage. E. muscae s.str. from Musca domestica is genetically very different from isolates originating from other naturally infected dipteran host taxa and correspondingly significant phenetic differences were observed (Jensen et al. 2001). Therefore we suggest that the nomenclaturally strict sense of E. muscae should be limited to isolates originating from M. domestica, the host from which it was originally described (Cohn 1855) and redescribed (Keller et al. 1999). However, it would be interesting to study the biogeographic variation of E. muscae s.str. using geographically diverse samples of the globally distributed M. domestica. E. muscae isolates from hosts other than M. domestica, such as Delia radicum and Scatophaga stercoraria, clustered together in the molecular analyses on separate branches that do not include the fungi from M. domestica. Other studies have also shown that each host fly species had its own distinctive fungal pathogen genotype, but that they fall within the same morphological parameter ranges as E. muscae s.str. (Jensen et al. 2001; Thomsen & Jensen 2002). This indicates that a radial
948
A. B. Jensen et al.
92
E. muscae s. str (Md)
E. muscae s. str. (Md)
74
100
100 E. muscae (Dr) 74
92
E. scatophagae (Ss) 95
Entomophthora sp. (Mym)
E. schizophorae (Md) E. schizophorae (Pr)
E. muscae (Dr)
E. muscae (Dr)
97
E. scatophagae (Ss) Entomophthora sp. (Mym)
98
E. syrphi s. str. (Mem) 100
E. muscae s. str. (Md)
E. scatophagae (Ss) Entomophthora sp.
99
E. syrphi s. str. (Mem) E. schizophorae (Md)
100 50 changes
E. schizophorae (Pr)
E. syrphi s. str. (Mem) E. schizophorae (Md)
100 50 changes
E. schizophorae (Pr)
Entomophaga 50 changes
A
B
C
Fig 1 – Phylogenetic relationships within the Entomophthora muscae species complex inferred from parsimony analyses of the first part of the LSU rRNA gene (A), the ITS II (B) and a combined data set (C). BS percentages over 50 % from 1000 replicates are shown above each supported branch. A. The single most parsimonious phylogram requiring 226 steps (CI [ 0.9336 and RI [ 0.8235), Entomophaga aulicae was used as outgroup. B. The single most parsimonious phylogram requiring 500 steps (CI [ 0.9420 and RI [ 0.9254), midpoint rooting was used. C. The single most parsimonious phylogram requiring 726 steps (CI [ 0.9439 and RI [ 0.9297), midpoint rooting was used. Md [ Musca domestica; Dr [ Delia radicum; Ss [ Scatophaga stercoraria; Mym [ Myospila meditabunda; Mem [ Melanostoma mellinum; Pr [ Pollenia rudis.
adaptation toward high host specificity may have occurred, and that these ecological differences can be used in the taxonomic separation of E. muscae s.l. Based on the molecular data two possible taxonomic solutions exist: either E. muscae should be maintained as one highly plastic species (which would require rejection of E. scatophaga as a separate species), or E. muscae (as currently recognized within the E. muscae species complex) should be segregated into multiple new species representing the distinctive genotypes affecting each of the host fly species or genera. We recommend the latter solution. E. schizophorae is easily recognized by both morphological and molecular criteria. A certain overlap between the natural host range for E. schizophorae and E. muscae exists, e.g., in host taxa such as M. domestica and D. radicum, while in other host taxa, such as Chamaepsila rosae, only E. schizophorae infections have so far been found (Eilenberg 2002; Eilenberg & Philipsen 1988). The sequence differences between the M. domestica and P. rudis isolates used in this study suggest a similar divergence, as seen within the E. muscae lineage. But until a thorough study of the variation of E. schizophorae from different host taxa is performed E. schizophorae should be recognized as a single species. E. syrphi can be distinguished molecularly from the other members of the E. muscae species complex. Morphologically, however, it can be difficult to differentiate some E. syrphi from E. muscae s.str. (Steenberg et al. 2001), but E. syrphi generally has larger conidia containing more nuclei. E. syrphi seems to have a restricted host range and has only been recorded from a few hover fly species (Ba1azy 1993; Keller 1987). Variable RFLP patterns have been observed within E. syrphi, suggesting the existence of yet another unexplored species complex (Thomsen & Jensen 2002). We were not able to cross-infect E. syrphi from hover flies to other fly species. However, in a study by Steenberg et al. (2001) an Entomophthora with a high number
of conidial nuclei and recognized as E. syrphi from a muscoid fly (Phaonia perdita) was shown to cause infection in M. domestica. An Entomophthora species with similar morphology but isolated from another muscoid host species (Myospila meditabunda) was included in the current molecular analysis, where it formed a monophyletic group together with E. syrphi. Nevertheless, we do not recommend including these Entomophthora species in E. syrphi because of their high sequence divergence and their different host ranges. E. syrphi, as well as E. grandis, should be restricted to Entomophthora pathogenic to hoverflies. Moreover the other species deserves to receive an amplified and more detailed species description. A species can be described as a single lineage of ancestor– descendent populations which maintains its identity from
Scatophaga stercoraria
Delia radicum
Coenosia Musca domestica tigrina
P. infirma B. fugax Neg. control 1500 bp 1200 bp 1000 bp 800 bp 700 bp 500 bp 400 bp 300 bp
Fig 2 – Agarose gel electrophoresis of RAPD amplifications obtained with the primer OPA-10. A molecular size ladder appears in the first and last lane. The Entomophthora muscae isolates obtained from a single host species had similar profiles, whereas isolates from different host species had different profiles. P. [ Pegoplata, B. [ Botanophila.
Entomophthora muscae species complex
other such lineages and which has its own evolutionary tendencies and historical fate (Wiley 1978). For species within the E. muscae complex, genetic characteristics, physiological host range and morphology of the conidia are entities that can be used to develop an operational taxonomic species concept. This study shows that the various regions of the ribosomal repeat, in particular the ITS II, include variation suitable for distinguishing species within the E. muscae species complex. However, in future other genomic regions should be included to perpetuate the recognition of phylogenetically defined species (Taylor et al. 2000). Physiological host range as evidenced by transmission experiments only provides a hint of the ecological host range; however, the correlation of molecular data with phenotypic evidence and host range will continue to improve our understanding of the evolutionary processes and species limits within the Entomophthorales.
Acknowledgements We wish to thank Verner Michelsen and Stig Andersen, Zoological Museum, Copenhagen for identification of fly hosts. From the Royal Veterinary and Agricultural University, Jan Martin is thanked for help with collection of Scatophaga stercoraria, Kirsten Ploug and Mette Vingaard are thanked for technical assistance. Nicolai Vitt Meyling for comments on the manuscript and Charlotte Nielsen for statistical support. The Danish Pest and Infestation Laboratory is thanked for providing flies and two in vivo cultures. Richard Humber, USDA, Ithaca and Ingeborg Klingen, Norwegian Crop Research Institute are thanked for in vitro cultures. The Danish Ministry of Food, Agriculture and Fisheries, the Royal Veterinary and Agricultural University and the Carlsberg Foundation supported this work financially.
references
Baird RB, 1957. Notes on a laboratory infection of Diptera caused by the fungus Empusa muscae Cohn. Canadian Entomology 89: 432–435. Ba1azy S, 1993. Flora of Poland, Fungi (Mycota), Vol. 24, Entomophthorales. Wydawnictwa Inst. Bot. Pan., Krako´w. Beauvais A, Latge´ JP, 1988. A simple medium for growing entomophthoralean protoplasts. Journal of Invertebrate Pathology 51: 175–178. Ben-Ze’ev IS, Kenneth RG, 1982. Features-criteria of taxonomic importance in the Entomophthorales. Part I. A revision of the Batkoan classification. Mycotaxon 14: 393–455. Cohn F, 1855. Empusa muscae und die Krankheit der Stubenfliege. Hedwigia 10: 57–61. Collet D, 1991. Modeling Binary Data. Chapman & Hall, London. Eilenberg J, 2002. Biology of fungi from the order Entomophthorales with emphasis on the genera Entomophthora, Strongwellsea and Eryniopsis: A contribution to insect pathology and biological control. D.Sc. Thesis. Department of Ecology, The Royal Veterinary and Agricultural University, Copenhagen. Eilenberg J, Bresciani J, Martin J, 1987. Entomophthora species with E. muscae-like primary spores on two new insect orders, Coleoptera and Hymenoptera. Nordic Journal of Botany 7: 577–584.
949
Eilenberg J, Philipsen H, 1988. The occurrence of Entomophthorales on the carrot flies (Psila rosae F.) in the field during two successive seasons. Entomophaga 33: 135–144. Freimoser FM, Jensen AB, Tuor U, Aebi M, Eilenberg J, 2001. Isolation and in vitro cultivation of the aphid pathogenic fungus Entomophthora planchoniana. Canadian Journal of Microbiology 47: 1082–1087. Higgins DG, Bleasby AJ, Fuchs R, 1992. CLUSTAL V: improved software for multiple sequence alignment. Cabios 8: 89–191. Humber RA, 1981. An alternative view of certain taxonomic criteria used in the Entomophthorales (Zygomycetes). Mycotaxon 13: 191–240. Humber RA, 1989. Synopsis of a revised classification for the Entomophthorales (Zygomycotina). Mycotaxon 34: 441–460. Jensen AB, Eilenberg J, 2001. Genetic variation within the insectpathogenic genus Entomophthora, focusing on the E. muscae complex, using PCR-RFLP of the ITS II and LSU rDNA. Mycological Research 105: 307–312. Jensen AB, Gargas A, Eilenberg J, Rosendahl S, 1998. Relationships of the insect-pathogenic order Entomophthorales (Zygomycota, Fungi) based on phylogenetic analyses of nuclear small subunit ribosomal DNA sequences (SSU rDNA). Fungal Genetics and Biolology 24: 325–334. Jensen AB, Thomsen L, Eilenberg J, 2001. Intraspecific variation and host specificity of Entomophthora muscae s.str. isolates revealed by random amplified polymorphic DNA, universal primed PCR, PCR-restriction fragment length polymorphism, and conidial morphology. Journal of Invertebrate Pathology 78: 251–259. Kalsbeek V, Mullens BA, Jespersen JB, 2001. Field studies of Entomophthora (Zygomycetes: Entomophthorales) d induced behavioural fever in Musca domestica (Diptera: Muscidae) in Denmark. Biological Control 21: 264–273. Keller S, 1984. Entomophthora muscae as a species complex. Mittelung Schweizer Entomologische Gesellschaft 57: 131–132. Keller S, 1987. Arthropod-pathogenic Entomophthorales of Switzerland. I. Conidiobolus, Entomophaga and Entomophthora. Sydowia 40: 122–167. Keller S, 2002. The genus Entomophthora (Zygomycetes, Entomophthorales) with a description of five new species. Sydowia 54: 157–197. Keller S, Kalsbeek V, Eilenberg J, 1999. Redescription of Entomophthora muscae (Cohn) Fresenius. Sydowia 51: 197–209. Kjøller R, Rosendahl S, 2000. Detection of arbuscular mycorrhizal fungi (Glomales) in roots by nested PCR and SSCP (single stranded conformation polymorphism). Plant and Soil 226: 189–196. Klingen I, Meadow R, Eilenberg J, 2000. Prevalence of fungal infections in adult Delia radicum and Delia floralis captured on the edge of a cabbage field. Entomologia Experimentalis et Applicata 97: 265–274. Kramer JP, Steinkraus DC, 1981. Culture of Entomophthora muscae in vivo and its infectivity for six species of muscoid flies. Mycopathology 76: 139–143. MacLeod DM, Mu¨ller-Ko¨gler E, Wilding N, 1976. Entomophthora species with E. muscae-like conidia. Mycologia 68: 1–29. Lo´pez Lastra CC, Hajek AE, Humber RA, 2001. Effects of two cryopreservation techniques on viability and pathogenicity of entomophthoralean fungi. Canadian Journal of Botany 79: 861–864. Mullens BA, 1989. Cross-transmission of Entomophthora muscae (Zygomycetes: Entomophthoraceae) among naturally infected muscoid fly (Diptera: Muscidae) hosts. Journal of Invertebrate Pathology 53: 272–275. Mullens BA, Rodriguez JL, Meyer JA, 1987. An epizootiological study of Entomophthora muscae in muscoid fly populations on Southern California poultry facilities, with emphasis on Musca domestica. Hilgardia 55: 1–41.
950
Nagahama T, Sato H, Shimazu M, Sugiyama J, 1995. Phylogenetic divergence of the entomophthoralean fungi: evidence from nuclear 18 S ribosomal RNA gene sequences. Mycologia 87: 203–209. Ramoska WA, Hajek AE, Ramos ME, Soper RS, 1988. Infection of grasshoppers (Orthoptera, Acrididae) by members of the Entomophaga grylli species complex (Zygomycetes, Entomophthorales). Journal of Invertebrate Pathology 52: 309–313. Remaudie`re G, Keller S, 1980. Re´vision syste´matique des genres d’Entomophthoraceae a` potentialite´ entomopathoge`ne. Mycotaxon 11: 323–338. SAS Institute Inc, 1999. SAS/STAT User’s Guide, 1st edn., Vol. 2, SAS Institute Inc., Cary, NC, USA, Version 8. Steenberg T, Jespersen JB, Jensen KMV, Nielsen BO, Humber RA, 2001. Entomopathogenic fungi in flies associated with pastured cattle in Denmark. Journal of Invertebrate Pathology 77: 186–197. Steinkraus DC, Geden CJ, Rutz DA, 1993. Prevalence of Entomophthora muscae (Cohn) Fresenius (Zygomycetes: Entomophthoraceae) in houseflies (Diptera: Muscidae) on dairy farms in New York and induction of epizootics. Biological Control 3: 93–100. Steinkraus DC, Kramer JP, 1987. Susceptibility of sixteen species of Diptera to the fungal pathogen Entomophthora muscae (Zygomycetes: Entomophthoraceae). Mycopathology 100: 55–63. Steinkraus DC, Kramer JP, 1988. Entomophthora scatophagae desc. ampl. (Zygomycetes: Entomophthorales), a fungal pathogen of the
A. B. Jensen et al.
yellow dung fly, Scatophaga stercoraria (Diptera: Anthomyiidae). Mycotaxon 32: 105–113. Swofford DL, 1998. PAUP: Phylogenetic Analysis Using Parsimony [Version 4]. Sinauer Assosiates, Sunderland, MA. Taylor JW, Jacobson DJ, Kroken S, Kasuga T, Geiser DM, Hibbett DS, Fisher MC, 2000. Phylogenetic species recognition and species concepts in fungi. Fungal Genetic and Biology 31: 21–32. Thomsen L, Jensen AB, 2002. Application of nested-PCR technique on resting spores from the Entomophthora muscae species complex: implications for analyses of host–pathogen population interactions. Mycologia 94: 794–802. Van de Peer Y, De Wachter R, 1994. TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Computer Applications in the Biosciences 10: 569–570. Watson DW, Petersen JJ, 1993. Seasonal activity of Entomophthora muscae (Zygomycetes: Entomophthorales) in Musca domestica L. (Diptera: Muscidae) with reference to temperature and relative humidity. Biological Control 3: 182–190. Wiley EO, 1978. The evolutionary species concept reconsidered. Systematic Zoology 27: 17–26. White TJ, Bruns TD, Lee SB, Taylor JW, 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JS, White TJ (eds), PCR Protocols: A Guide To Methods And Applications. Academic Press, San Diego, pp. 315–321.